The best HMI screens are boring until something goes wrong — then they are impossible to misread. Here is how experienced SCADA engineers get there, and the mistakes that make operators stop looking.
Most HMI screens are not designed — they accumulate. A pump gets added here, a flashing red banner there, a photo-realistic 3D tank because the integrator's library had one. Two years later the overview screen has forty animated widgets, six shades of red, and operators who have learned to ignore all of it. That is the real cost of bad HMI design: not ugliness, but alarm blindness and slow abnormal-situation response.
The industry answer to this is usually called high-performance HMI (popularized by the ISA-101 standard and the PAS/Hollifield work): gray backgrounds, minimal color, and screens organized around what the operator must decide, not around what the equipment looks like. You do not need to adopt every rule to benefit — the practices below are the subset that pays off on every project, from a two-pump skid to a plant-wide SCADA system.
Muted background, gray equipment, neutral values. Red means critical alarm, amber means warning — and nothing else on the screen is ever red or amber. Color that always means something is color operators trust.
One screen should answer "is my area healthy?" in five seconds: key process values, active alarm counts, equipment run states. Details live one click deeper, not on the overview.
A number without its normal range forces the operator to remember 400 setpoints. Add units, expected range, and a simple trend or gauge so abnormal is visible at a glance.
Overview → area → equipment detail. Three levels, consistent placement, same navigation buttons on every screen. If an operator needs a map to find a valve, the hierarchy failed.
An alarm the operator cannot act on is noise. Every configured alarm should have a defined operator response — if there is not one, it should be an event log entry, not an alarm.
Screens are judged during upsets, not steady state. Simulate alarm floods and failure scenarios before commissioning and watch whether the screen still communicates priority.
Do not wait for hardware. Drive the HMI from a Modbus simulator playing realistic values and failure patterns, and iterate on the design while changes are still cheap.
If you standardize only one thing, standardize color. A workable scheme used across thousands of plants:
| Color | Meaning | Allowed anywhere else? |
|---|---|---|
| Red | Critical alarm — immediate action | No. Not for "stopped", not for logos, not for buttons. |
| Amber / Yellow | Warning — attention soon | No. |
| Blue / White | Advisory, information | Sparingly. |
| Gray tones | Normal equipment and background | Everywhere — normal should look calm. |
| Green | Running / healthy (use with restraint) | Avoid large green areas; a screen full of green hides the one thing that is not. |
The test is simple: put the screen on a wall, stand back three meters, and trigger one critical alarm. If your eye does not land on it inside one second, the screen has too much competing color.
Gauges vs numbers: analog-style gauges and bar indicators communicate "where in the range am I?" faster than digits. Use digits when the operator needs the exact value (setpoints, totalizers), gauges when they need the trend of health (pressures, levels, temperatures).
Trends on the screen, not behind a click: a small sparkline next to a critical value answers "is it rising?" without navigation. Full trend analysis can live in a dedicated historian view, but direction-of-travel belongs on the operating screen.
Consistent equipment symbols: a pump looks like the same pump on every screen, rotated the same way, labelled the same way. Symbol libraries help, but discipline helps more — the goal is that operators read state, not decode artwork.
Whitespace is functional: group related equipment visually and let unrelated groups breathe. Density feels productive to engineers and is hostile to operators at 3 a.m.
ISA-101 is the standard covering human-machine interface design for process automation: display hierarchy, color usage, navigation, and alarm presentation. You do not need formal compliance to benefit from its core ideas — most of this article is aligned with it.
Dim control rooms favor dark themes (less glare); bright plant floors favor light themes. Modern SCADA tools let operators toggle — what must not change between themes is the meaning of alarm colors.
Run a Modbus simulator as the data source and build screens against realistic changing values. You can rehearse alarm floods, sensor drift, and equipment failures safely — see our guide to building an HMI without panel software.
Common guidance targets fewer than ~150 alarms per operator per day and no more than ~10 in any 10-minute flood. If you are far above that, the fix is alarm rationalization, not a bigger alarm list.
Our SCADA includes a drag-and-drop HMI designer with gauges, tanks, pumps, alarms and trends — and you can drive it from a simulated or real Modbus device from day one. Desktop from $199 perpetual, or SCADA Cloud free for 30 days.